Volume 44, Issue 2, Pages 263-277 (October 2004) Ndel1 Operates in a Common Pathway with LIS1 and Cytoplasmic Dynein to Regulate Cortical Neuronal Positioning  Tianzhi Shu, Ramses Ayala, Minh-Dang Nguyen, Zhigang Xie, Joseph G. Gleeson, Li-Huei Tsai  Neuron  Volume 44, Issue 2, Pages 263-277 (October 2004) DOI: 10.1016/j.neuron.2004.09.030

Figure 1 Ndel1 RNAi Specifically Silences Ndel1 Expression (A) In NIH3T3 cells, endogenous Ndel1 expression is significantly silenced by Ndel1 RNAi (60% ± 6%, mean ± SEM, Student's t test, p < 0.01). It is noted that the number does not indicate the reduced protein level in individual cells because of the compromise of the DNA transfection efficiency. Proteins associated with Ndel1, such as LIS1, DIC, p50 (dynactin), and p150glued are not affected. Actin indicates an equal loading. (B) In individual 3T3 cells cotransfected with EGFP (green) and control or Ndel1 RNAi (white arrows), Ndel1 RNAi silences Ndel1 expression (cy3, red) while LIS1 expression is not altered (cy5, blue). Control RNAi has no impact on the expression of Ndel1 or LIS1. Scale bar, 20 μm. (C) Ndel1 expression is significantly silenced in E15 cortical neuronal culture after the electroporation of Ndel1 RNAi (62.5% ± 13%, Student's t test, p < 0.01). LIS1 and DHC level are not changed. (D) In cortical neurons cotransfected with EGFP (green) and control or Ndel1 RNAi (white arrows), Ndel1 RNAi silences Ndel1 expression (red) without affecting LIS1 (blue). Control RNAi does not silence Ndel1. Scale bar, 20 μm. (E) The mutated Ndel1 sequence targeted by RNAi (the red letters indicate three point mutations). Transfection of the wild-type and the mutant Ndel1 results in augmented expression of Ndel1 protein. Ndel1 RNAi silences the overexpressed Ndel1 encoded by the wild-type but not the mutant Ndel1. Neuron 2004 44, 263-277DOI: (10.1016/j.neuron.2004.09.030)

Figure 2 Ndel1 Is Required for Dynein Activity (A) The experimental paradigm of the microtubule transport assay. (Aa) Nucleation of the fresh microtubules from the centrosome after nocodazole was washed off. (Ab) Segments of microtubules are transported to the cell periphery. (B) In control RNAi and EGFP cotransfected cells, microtubules (red) were transported from the perinuclear region to the cell periphery. In cells treated with DHC, LIS1, and Ndel1 RNAi, DHC, LIS1, and Ndel1 expression are significantly silenced (cy5, blue). Microtubules accumulated in the perinuclear region (red). The green circle demarcates the perinuclear region. The green dotted line traces the cell membrane, and the area between the green line and green circle is defined as the peripheral region. The fluorescence background within the nucleus was subtracted from all measurements. (C) Quantification of the distribution of the microtubules in the perinuclear versus the cell peripheral region. DHC, LIS1, and Ndel1 RNAi caused a significant increase of this ratio (Student's t test, p < 0.0001). Error bars indicate SEM. Scale bar in (B), 10 μm in all panels. Neuron 2004 44, 263-277DOI: (10.1016/j.neuron.2004.09.030)

Figure 3 Ndel1 Facilitates the Interaction of LIS1 and Dynein (A) Immunoprecipitation from lyses of N2A cells treated with control or Ndel1 RNAi. Less LIS1 and Ndel1 are coimmunoprecipitated by anti-DIC antibody in Ndel1 RNAi treated cells. (B) Quantification of coimmunoprecipitation experiments. Ndel1 RNAi significantly silenced Ndel1 expression (Student's t test, p < 0.01). LIS1 and Ndel1 associated with DIC were significantly reduced (Student's t test, p < 0.05). (C) Glycerol gradient from control or Ndel1 RNAi treated cells. In the cells treated with Ndel1 RNAi, the presence of LIS1 in the same fractions with DIC (the boxed region) is significantly reduced, although the overall distributions of DIC or LIS1 are not changed. (D) Quantification of the glycerol gradient experiments. Followed by Ndel1 RNAi silencing, the amount of LIS1 level that coexists with DIC was significantly reduced (Student's t test, p < 0.05). Error bars indicate SEM. Neuron 2004 44, 263-277DOI: (10.1016/j.neuron.2004.09.030)

Figure 4 Ndel1, LIS1, and DHC RNAi Impair Nuclear Movement and Disrupt Microtubule Bundles Coupling the Nucleus to Centrosome in Cortical Neuron Cultures (A) Two examples of cortical neurons cotransfected with EGFP and control RNAi. These cells display a typical nuclear translocation with the nuclei (indicated by the arrows) moving consistently toward one prominent process. (B) Two examples of cortical neurons cotransfected with EGFP and Ndel1 RNAi. These cells display abnormal nuclear movements. The nuclei moved back and forth without a determined direction (oscillation) and repeatedly shrunk and expanded without obvious translocation. Frequently the nuclei simultaneously display both abnormalities. (C) Ndel1, LIS1, and DHC RNAi significantly decrease the distance of the nuclear movement. (D) Ndel1, LIS1, and DHC RNAi significantly increase the percentage of neurons displaying nuclear oscillation and contraction. (E and F) Transfection of Ndel1, LIS1, or DHC RNAi in cortical neurons disrupts microtubule bundles that couple nucleus to centrosome (white arrows indicate microtubule bundles that enwrap the nucleus). (G and H) Transfection of Ndel1, LIS1, and DHC RNAi cause the uncoupling of centrosome and nucleus (arrows indicate the centrosome labeled with RFP-CentrinII). Error bars indicate SEM. *p < 0.05, **p < 0.01. Scale bar in (A): 10 μm in all panels. Scale bar in (B): 4 μm in panels with high magnification and 10 μm in panels with low magnification. Scale bar in (E) and (G): 5 μm for all panels. Neuron 2004 44, 263-277DOI: (10.1016/j.neuron.2004.09.030)

Figure 5 Silencing of Ndel1 In Utero by RNAi Electroporation Causes Neuronal Migration Defect (A) Schematic of in utero RNAi electroporation. (B) Embryos were coelectroporated with Ndel1 RNAi and EGFP on E15 and perfused on E17. Coronal brain sections were immunostained with Ndel1 antibody. Ndel1 expression levels are reduced in the cortical neurons at the junction of the SVZ and the IZ (arrow). High-magnification panel shows that Ndel1 expression is silenced in two electroporated neurons (arrows). (C) Coronal section of E17 brain coelectroporated with EGFP and control or Ndel1 RNAi vectors on E15. Coelectroporated neurons are shown in green. Sections are counterstained with DAPI (blue). (D) Quantification of the neuronal positioning in control and Ndel1 RNAi electroporated brains. All electroporated neurons (EGFP positive) across the whole cerebral cortex were counted. Bar graph shows percentage of the total electroporated neurons distributed in SVZ/VZ, IZ, and CP. Error bars indicate SD. ** indicates differs from control at p < 0.01. (E) Electroporated neurons in the SVZ/VZ were colabeled with BrdU 24 hr after RNAi electroporation (red). There is no difference in the percentage of BrdU-positive neurons between control and Ndel1 RNAi treated brains (Student's t test, p = 0.507). The inset shows two colabeled neurons. Scale bar in (B): 200 μm in the three low-magnification panels in (B) and (C); 20 μm in the high-magnification panel in (B); and 40 μm in (E). Neuron 2004 44, 263-277DOI: (10.1016/j.neuron.2004.09.030)

Figure 6 Ndel1 Operates with LIS1 and Dynein in a Common Pathway during Cortical Neuronal Migration (A and B) Coronal sections of P4 brains that were coelectroporated on E17 with the indicated constructs in conjunction with EGFP. The rostral regions of the brain are illustrated in two separate panels. The top panels show the frontal cortex, and the middle panels show the SVZ/VZ. All cortical layers in the caudal cortex are illustrated in the bottom panels. Brain sections are counterstained with DAPI (blue). The asterisk indicates the midline of the forebrain. CC, corpus callosum. (C) Quantification of the in utero RNAi electroporation experiments. Neurons distributed in SVZ, IZ, and CP were counted through the whole brain. Error bars indicate SD. *p < 0.05, **p < 0.01. Scale bar in (B): 200 μm in all panels. Neuron 2004 44, 263-277DOI: (10.1016/j.neuron.2004.09.030)

Figure 7 Loss of function of Ndel1, LIS1, and DHC Uncouples the Centrosome and Nucleus In Vivo E15 embryos were electroporated with EGFP/RFP-CentrinII/control, Ndel1, LIS1, or DHC RNAi. Embryos were collected and perfused on E17.5 for confocal analysis. RFP-CentrinII labeled the centrosome in cortical neurons, and the nuclei were stained with DAPI. Neurons that migrate from the SVZ to the IZ display a stereotypical migratory morphology with a leading process (arrows indicate such neurons in [A]–[D], and the high-power images of these neurons are also shown). In neurons electroporated with control RNAi, the centrosome is positioned closely to the nucleus on the side of the leading process (arrowhead in [Aa1] indicates the centrosome, and the white lines demarcate the nuclear boundary). In neurons electroporated with Ndel1, LIS1, or DHC RNAi (B–D), this distance is significantly increased. Scale bar in (Dd1): 10 μm in (Aa1)–(Dd1); 60 μm in (A)–(D). (E) The quantification of the effect of loss of function of Ndel1, LIS1, and DHC on the coupling of the centrosome and nucleus in cortical neurons. Error bars indicate SEM. *p < 0.05. Neuron 2004 44, 263-277DOI: (10.1016/j.neuron.2004.09.030)

Figure 8 A Model of How Ndel1 Operates with LIS1 and Dynein to Regulate Neuronal Migration (A) In the physiological condition, Ndel1 facilitates the interaction of LIS1 and dynein and the formation of the Ndel1/LIS1/dynein complex. This complex is important for the integrity of the microtubule bundles that couple nucleus and centrosome, which mediates proper nuclear translocation that is essential for neuronal migration. When Ndel1 is silenced (B), the Ndel1/LIS1/dynein complex disassembles (silencing LIS1 or DHC presumably causes the disassembling of this complex as well). The microtubule bundles coupling nucleus and centrosome are disrupted, which causes abnormal nuclear translocation and leads to the deficit of neuronal migration and positioning in vivo. The microtubule structure in the leading process is not significantly disrupted, possibly due to other plus end binding proteins such as Clip family, EB1, and APC. Neuron 2004 44, 263-277DOI: (10.1016/j.neuron.2004.09.030)